e/local-llm 🚀 Working on scaling local inference LLM fine tuning | Harness optimization | Evals and Memory systems for Agents (opinions my own)
Tokyo
Joined September 2016
- Tweets22K
- Following575
- Followers21.4K
- Likes50.7K
Pinned Tweet
This tweet is about how I have studied ML and made it my profession. I'll share the resources I've used and the sequence of my study.
Straight to point
ML pre-requisites(maths) : Linear Algebra, Probability Theory, Calculus, Optimization Theory(optional), Information theory(optional)
Linear Algebra: Lecture course by Gilbert Strang
Probability theory: MIT 6.041 (it contains parts of Bayesian inference as well)
Calculus: your high school and college classes are enough
Once basic maths is done then we move to ML.
Classical ML : CS229. Either by Andrew NG or someone else. Follow their lecture notes and solve their problem sets.
Reference books for classical ML that I followed: PRML by Christopher bishop, Pattern Classification by Duda, Hart and Stork
After getting comfortable with classical ML we move to Deep Learning and everything else.
Deep Learning and Computer Vision: CS231n. Very good lecture and assignments
Reference book: Deep Learning by Ian Goodfellow. This is the best book on deep learning. I’ve read some chapters of it many many times. Beautiful maths and intuitions
MLOps: dvc, WandB, MLFlow
NLP: I just read hugging face blogs. I haven’t spent much time with classical NLP though.
Alignment/AI safety/AI explainability: Anthropic Blogs(I’m a noob in this, just started learning couple months ago)
Additionally:
Blogs: Lilian Weng(OpenAI)’s blogs, colah’s blogs
Additionally: arxiv. I read many papers from arxiv
Karas and Tensorflow blogs: for introductory code about modern deep learning frameworks
Competitions: Kaggle
Cloud compute. GCP/collab/Kaggle notebooks
PS: this is not a roadmap. Just what I followed till now and I find it quite structured. Even after 5 years I still find myself learning new stuff everyday.
> GMO is urging users to unplug network cables,
Yabai desu
research isn’t a gift by one company to anyone else
Sure “Attention is all you need” got popular but very similar ideas were floating around that time. In fact the concept of softmax attention wasn’t new in 2016/17
researchers build the foundation ground up sharing.
Saurabh Kumar retweeted
Currently, browsers and agents live separately, which doesn't make any sense.
Strawberry puts your agents precisely where the work happens. Less friction, less heckling, and more autonomy.
The first browser that's built for agents at scale! give it a shot
Currently, browsers and agents live separately, which doesn't make any sense.
Strawberry puts your agents precisely where the work happens. Less friction, less heckling, and more autonomy.
The first browser that's built for agents at scale! give it a shot
Saurabh Kumar retweeted
Always the same “joke” - it never seems to get old to them. In case you forgot, smell—and the way we talk about it—is deeply political.
Saurabh Kumar retweeted
No one sees more startups than I do - all around the world. Spent three weeks in India with 15 companies. Quite convinced about what is going on. Never been to a place where the VCs have a worse reputation.
Saurabh Kumar retweeted
Is "Indian nepotism" real? Maybe.
Immigrant-founded firms tend to employ a lot of coethnics—especially for groups that don't really know English. Chinese founders fill 45% of first hires with Chinese, Mexicans 31%, and Poles 27%.
Indians are at 21%, below the immigrant average.
How do we build trust ladder for memory?
AI writes to slack, to notion, to JIRA and then reads back from it, considering it the source of truth?
One misleading fact and it becomes impossible to spot it unless it's too late
Coding on the other can provide immediate feedback
Saurabh Kumar retweeted
Once again, communication in Japanese is hell if you are autistic.
Everything is secretly a code for something else. Nothing is ever said directly - even if the meaning is immediately apparent to everyone "in the know".
Saurabh Kumar retweeted
Tunnel Boring Machines का पूरा ecosystem भारत में बन रहा है।
Design ➡️ Manufacturing ➡️ Maintenance
आत्मनिर्भर भारत 🇮🇳
Saurabh Kumar retweeted
Being fearless, not timid; being smart, not slimy; being gentle, not rude; and being skillful, not callous. These are all signs of the Daivi Shakti arising in you, and that is the celebration of Dussehra.
#Vijayadashami
#Dussehra
>India isn't developed, it's developing, so it needs patriots, not parasites
Preach
I think what is actually needed is for Indian people to actually give a damn about their cities and fellow citizens.
Anybody using the "I pay taxes, I shouldn't have to do anything" excuse should go live in another country. India isn't developed, it's developing, so it needs patriots, not parasites. If this triggers you, it means you're part of the problem.
If you're not happy with the footpath (or a pothole, or anything else which is broken), fix it, or call someone to fix it, and don't give up until it's fixed. Be relentless. Be vocal. Be creative. Make it your mission to get that one thing fixed. Look at what Virendra (@Virendra0698) did by just calling the right people and being respectful (but also persistent). He and Arun Pai (@arunpaiblrwalks) along with the help of GBA and some wonderful volunteers turned an unofficial garbage dump into a pristine, beautiful footpath through sheer force of will.
xitter.cf/Virendra0698/status/21…